MovieSeer was my submission for HackPrinceton 2012. It was a website that would recommend movies for you to watch. The recommendations were generated using your ratings for movies you've already seen. It used the Rotten Tomatoes API to get movie metadata and would try to find movies that are similar to ones you like. So it was a very simplistic content-based filtering approach.
It won the "Best Use of MongoDB" award. I still have the check that I got for it, though I believe Mongo got rid of the monetary prize the next time they sponsored the hackathon, so they probably figured it wasn't a great investment.
Note that the Rotten Tomatoes API endpoint that I used is no longer available. There's still an API, but it requires an application for access, and it doesn't provide movie metadata.
I created my GitHub account for this hackathon.
I volunteered to be the first person to present my project for judging. During
my live demo, I hit the button to get recommendations, and...it hung for
several seconds (though it felt like an eternity). The results eventually
appeared. I later realized I had forgotten to remove a print
statement that
ran in a loop. If you look at the commit history, there's a commit with a
"demo version (need to remove print statements)" messsage. Lesson learned!
MIT